A simple iterative algorithm based on least-squares minimization was designed to precisely localize particles based on fluorescence images. The algorithm is based on least-squares minimization for the case where the errors in pixel values are independent of the number of counts in the pixels. In this situation the general non-linear minimization problem simplifies to iterating an x-y location until convergence (with generally less than five iterations). Once the center of the particle has been found, the integrated intensity is then simple to compute. This algorithm improves on a previous version with roughly twice x-y precision and much improved integrated intensity precision.